Project Background/Rationale. Parkinson’s disease (PD) is the second most common neurodegenerative disease, affecting over one million Americans. The cardinal clinical manifestations of PD include resting tremor, rigidity, bradykinesia, and postural instability/gait disturbance (Olanow et al., 2009). In addition, people with PD fall frequently, with 60% falling annually and two-thirds of these falling recurrently (Latt et al., 2009; Allen et al., 2013; Bloem et al., 2001). Identifying interventions that successfully improve postural control and reduce fall rate is critical to reduce disability, improve quality of life, and potentially increase survival in patients with PD. Recent randomized, controlled trials (RCT) have examined the effects of exercise and physical therapy interventions on reducing falls in patients with PD; however, with mixed results. Project Objectives. There is a limited availability of effective treatment options to reduce falls in PD. In this context, evidence from two meta-analyses (Sherrington et al., 2017; Allen et al., 2011) indicates that highly challenging exercise approaches may lead to better outcomes. We propose investigating the effects of two theoretically driven, progressive, highly challenging exercise programs: 1) a facility-based in-person exercise program that demonstrated a significant decline in fall rate over time in a small sample of patients (Sparrow et al., 2016) and 2) a novel home-based program centered around remote, real-time exercise instruction and supervision via regular videoconferencing. Both programs will be delivered by a physical therapist with experience treating patients with PD. There will also be a control group in which health education is provided. Project Methods. We propose to conduct an RCT evaluating effects on fall rate, dynamic balance, and fear of falling. A total of 162 VA patients with mild-to-moderate PD will be randomly assigned to one of the three 3- month interventions: in-person exercise at the VA, remotely-delivered exercise, or health education. Outcomes will be compared between each intervention group and the control group. The primary outcome will be the fall rate. Fall rates will be compared between groups with the use of negative binomial regression models.